b'determining the potential influencethe high- the transfer efficiency by 32% over theair pressure (and flow) caused the of the boundary layer flow on thevelocity spraybaseline case. The larger improvementlargest reducing in volume transfer. spray trajectory for two reasons. First,leaves theof the knife-edge setup relative toAs was observed in the boundary it is an integral quantity calculatedPIV regionthe blunt plate setup is likely due tolayer thickness and momentum across the boundary alter and isat a higherthe upstream-facing knife-edge platethickness results, causing the blow-therefore more robust than the directposition,having its downstream-edge closer tooff nozzle setup to i) increase the boundary layer thickness. Second,which wouldthe spray nozzle, resulting in a moreboundary layer thickness allowing the and more importantly, the momentumlikely lead tocomplete blockage of airflow aroundair-barrier to interact with the spray thickness provides an assessmentan increasedthe spray, whereas the blunt platefor a longer time and over a larger of the effective momentum state ofin lost spraysetup leaves a larger gap for entrainedspace, increasing the altered spray the boundary layer, and, with thisfluid air to backfill and begin disrupting thetrajectory away from the web surface. investigations focus on the abilityspray further away from the plate. ThisAnd ii) the blow-off nozzle actually of the boundary layer to redirect theanalysis is un-confirmed since thisadded high-momentum air flow to the spray droplets, momentum (i.e. drag)region was too far from the plate to beboundary layer, causing an increase is an attractive and possibly morecaptured in the PIV data field of view. in high-momentum fluid as can be representative metric. As can be seenseen in Figure 8. This increase moved in Figure 5, the arrangement fromThe rod setup had a notable influencethe bulk spray trajectory downstream, high to low momentum thicknessnear the rod, but did not have aas can be seen in the spray-region of in the spray region actually followspersistent effect downstream at thethe PIV results when comparing the what was observed for the boundarylocation of the spray. This is reflectedbaseline data (Figure 11) to the blow-layer thickness. However, the resultsin the 3% change from the baselineoff data (Figure 13). Also, it appears are smoother and amplified in their a negligible amount. Finally, thefrom the most downstream portion difference from the baseline (black)blow-off setups had the worst effectof Figure 13, that the high-velocity case, where =3.0; with the blow-offin spray transfer efficiency as theyspray leaves the PIV region at a higher case increasing by 123% from theactually reduce the transfer efficiencyposition, which would likely lead to an baseline and the knife-edge at theby 16 and 26%, where the higherincreased in lost spray fluid.low-position decreasing by 53% from the baseline. Setup Transfer Efficiency Change from baseline(%)The results from the fluid collection testing are provided in Table 1. TheseKnife -Edge (0.25" gap) 45% 45%results agree with the investigatedKnife -Edge (0.125" gap) 44% 42%trends from the PIV analysis very well. The knife-edge results increasedBlunt-Plate (0.25" gap) 41% 32%the transfer efficiency the most by 42-45% depending on the gapRod (0.25" gap) 32% 3%distance. The knife-edge setupBaseline 31% N/Aprovides an interesting arrangement by moving the boundary layer airRod (0.125" gap) 30% -3%away from the wall, forcing a back-fill behind the plate resulting inBlowoff (25 psi) 26% -16%reversed flow and a further decreaseBlowoff (45 psi) 23% -26%in boundary layer velocity. The blunt plate also performed well, increasingTable 1: Volume collection and transfer efficiency results.WORLD PULP&PAPER 79'